Are Cat7 Cables Faster Than Cat5e?
We Needed New Cables, So Why Not Upgrade?
With the COVID-19 virus requiring me to work from home, I decided to rearrange my home office to be more conducive to productivity. Part of the move was running cables under the floor so that the Eufy robovac wouldn't get hung up on them. To accomplish this, I needed some newer cables, longer than anything I had laying around, so I decided to order some new ones from Amazon. That's when I discovered the existence of Cat6 and Cat7 cables (and even saw some references to Cat8). After a little research, I learned that Cat7 cables are rated for higher speeds than Cat5 and Cat5e. The cost was fairly comparable, so I ordered Cat7.
The Cable Ratings Explained
The primary difference between cat5 (including cat5e), cat6, and cat7 is the amount and type of shielding. In short, the newer offerings have greater shielding against interference from outside sources, cross-talk from one pair of wires to another, and data loss. This additional shielding allows the products to rate the theoretical data transfer speeds higher. Ratings are in both throughput (megabits per second or Mbps) and frequency (megahertz or MHz). The frequency is just how fast the data on a pair of wires can change from one state to another.
The Test Parameters
To conduct these tests, I started with my old setup: my trusty Dell Latitude 5580 laptop, a 10-year-old HP USB Hub, a TrendNet 24-port Gigabit Switch, a Netgear R6350 dual-band router, and an Arris cable modem provided by my cable company. The laptop connected to the USB Hub with the provided USB cable. This HUB connected to the switch, the router to the switch, and the router to the cable modem using generic cat5e cables I constructed myself several years ago. All of these cables were six feet in length. The first round of tests comprised three executions of the speed test from speedtest.net, which gives you the ping response time in milliseconds, and the download and upload speeds in megabits per second.
Round 2, USB Hub Removed
For round two of testing, the only change was to remove the USB hub from the configuration, and connect the switch directly to the laptop using the cat5e cable. The results showed a significant increase in download speed, indicating that the USB hub could be a bottleneck.
Round 3 With New Cat7 Cables
For the third round, I replaced all the Cat5e cables with Cat7 cables. All of the cat5e cables were replaced with 10-foot Cat7 cables except for the one from the switch to the laptop. This is the one that will eventually be run under the floor, so the new cable is 20 feet long. So was there a significant increase in any of the measured results? In a word, no.
Last Round, Re-add the USB Hub
For the fourth and final round, I re-inserted the USB hub between the laptop and the 24-port switch. As the results confirm, the hub does indeed negatively impact download speed.
Summing Up the Ethernet Cable Test Findings
So, at first blush we might find the fact that the new cables did not result in better upload or download speeds surprising, but there are actually good reasons for that. First of all, the service from our cable company probably limit us from seeing any results much better than 115 Mbps in any case. It is quite possible that if our service provided speeds approaching gigabits per second, the Cat7 cables might have been able to handle them far better than the old Cat5 ones. Also, if the environment had a lot of electronic "noise" in the form of strong wireless signals or interference from high tech devices, the signals in the Cat5 cables might suffer from measurable degradation that the Cat7 cables protect against. Finally, the relatively small number of runs over a very limited period of time doesn't give us a real robust picture. It's quite possible that the load on our local cable company's infrastructure or the demand on the speedtest.net servers could significantly impact the results. The tests are of some value, however, in showing that for normal home or small business networks, all that extra shielding is probably not going to improve throughput very much, if at all.